The author is learning Flux, a new language developed at InfluxData, to analyze cryptocurrency exchange data. They start with a basic query using the `from`, `range`, `filter`, and `window` functions to extract specific fields from the "crypto/autogen" bucket over the past 48 hours for Bitcoin (USDT_BTC). The results are overwhelming, so they filter out unnecessary data and focus on just Bitcoin. They then apply a window function to group the data by hour and calculate the mean of each field. Finally, they use the `group` function to combine the relevant values into fewer tables. The author finds the Flux query readable and easy to understand, with only one dislike - understanding the table structure being returned. Overall, they successfully write a useful Flux query that allows them to make meaningful comparisons in their data.